Sentence starters: 227 examples for essays and academic writing

Sentence starters are the words that open a sentence and tell the reader what it is about to do: add a point (In addition), give an example (For instance), push back (However) or wrap up (In short). The 227 below are grouped by that job, so you can pick the one that matches what your sentence actually does rather than the one that sounds most formal. They are written for essays and research papers first, and most of them work just as well in reports, articles and long emails.

This list works one sentence at a time. The first sentence of a paragraph has a bigger job, because it also has to state what the paragraph will prove, so treat a starter there as the first few words of a topic sentence, not the whole of it. Every starter below comes with a full example sentence, so you can see the punctuation that goes with it. Find the job in the table, choose a starter, then read your sentence once without it. If the logic is still clear, you may not need a starter at all.

Starters to introduce a topic

For the first sentences of an essay, a chapter or a section, after any opening hook. For the very first line of an essay, see essay hook examples.

1. This essay examines how the 1918 influenza pandemic changed public health campaigns in American cities.
2. The purpose of this study is to measure how often students revise their first drafts.
3. This report sets out the costs and benefits of moving the archive to a new building.
4. The following analysis compares two approaches to teaching fractions in the first year of middle school.
5. At the center of this debate is a simple question: who owns a photograph of a public building?
6. Few questions in urban planning divide residents as sharply as where to put the parking.
7. This chapter turns from the causes of the strike to its consequences.
8. The question this paper addresses is whether bilingual children learn to read later than monolingual ones.
9. This section introduces the three case studies the rest of the paper relies on.
10. The focus here is narrower than the title suggests: one river, one decade and one set of dams.
11. Recent debates about remote work have largely overlooked one group: new employees.
12. To understand the reforms, one first has to understand who could vote before them.
13. The aim of this review is to bring together studies of reading on paper and on screens.

Starters for background and context

For the sentences that tell the reader what they need to know before your argument makes sense: history, definitions, earlier research.

14. Historically, the city drew all of its water from a single reservoir north of town.
15. For most of the twentieth century, the term was used only by specialists.
16. Before the redesign, the application form asked for just two pieces of information.
17. Traditionally, a thesis opens with an introduction and a review of the literature.
18. In recent years, more museums have put their collections online.
19. Over the past decade, the number of applicants to the program has nearly doubled.
20. At the time the survey was designed, few households owned a smartphone.
21. Researchers have long assumed that the two species never share a habitat.
22. Some background on the dispute is needed here, because both sides use the same terms differently.
23. The term needs a definition, since it means different things in law and in economics.
24. Most of the existing research has focused on adults rather than teenagers.
25. As the previous chapter showed, the reforms began in the ports, not the capital.
26. Earlier studies relied on surveys that students filled in at home.
27. The company began in 1998 as a two-person consultancy above a hardware store.
28. A brief history helps explain why the rule is still on the books.

Starters to state a claim

For the thesis and the topic sentences that carry it. A claim should be something a reader could disagree with; see thesis statement examples for whole sentences built this way.

29. This essay argues that the narrator's silence in the final chapter is a choice, not a failure.
30. I will argue that the policy failed because no one measured what it was meant to change.
31. The central claim of this paper is that class size matters most in the first two years of school.
32. The evidence points to a single conclusion: the bridge was underdesigned from the start.
33. Contrary to the usual reading, the poem is about work, not love.
34. What this case shows is that rules without enforcement become suggestions.
35. The real problem is not the cost of the program but its timing.
36. It is reasonable to conclude that the second draft is the one the author meant to publish.
37. The strongest explanation for the decline is the closure of the rail line in 1964.
38. My position is that the museum should return the manuscript to the archive it came from.
39. Rather than a single cause, the fire had three, and each made the next one worse.
40. This paper makes two claims: first, that the data were collected too early; second, that the sample was too small.
41. The novel's ending, I will suggest, undoes everything its first chapter promised.
42. Read closely, the letters tell a different story from the official report.

Starters to add a point

For the second and third reasons in an argument. Vary them: a run of sentences that open with Furthermore reads like a list someone forgot to format.

43. In addition, the second survey reached twice as many households.
44. Furthermore, the interviews revealed a problem the survey had missed.
45. Moreover, the cost would fall on the people least able to pay it.
46. Another reason to doubt the account is that no other witness mentions the storm.
47. A second factor is the weather during the harvest.
48. Beyond the financial cost, the delay damaged trust between the two teams.
49. Equally important is the question of who will maintain the system.
50. Not only did the reform cut costs, but it also shortened waiting times.
51. On top of this, the original plans had been lost.
52. Along the same lines, the second letter repeats the request almost word for word.
53. What is more, the author revised the passage three times.
54. Added to this is the problem of translation.
55. Further evidence comes from the parish records.
56. Besides the legal risk, there is a practical one: nobody knows where the original files are.
57. A related point concerns the timing of the announcement.

Starters to give an example

For the sentence that turns a general point into a case the reader can picture. The example usually works best right after the claim it supports.

58. For example, the committee's first report lists only one woman among its forty contributors.
59. For instance, a student who misses the first week has to learn the course's vocabulary alone.
60. To illustrate, consider a reader who opens the book at chapter ten.
61. A case in point is the town of Millbrook, which kept its station open when its neighbors lost theirs.
62. Take, for example, the first chapter of Pride and Prejudice, which is mostly dialogue.
63. Consider the case of a small bakery that raises its prices by ten percent.
64. This is clearest in the second interview, where the manager contradicts herself twice.
65. In particular, the report criticizes the lack of training.
66. Specifically, the error appears in every table after the first.
67. One such case is the 1927 Mississippi flood, which drove hundreds of thousands of people from their homes.
68. Examples include the river project, the school buildings and the new hospital wing.
69. This can be seen in the way the characters speak to their servants.

Starters to cite a source

For bringing in research, data, documents and quotations. The verb matters: argues, found, suggests and claims each tell the reader how much weight you give the source. For how citations fit into a full paper, see the research paper example.

70. According to Okafor (2022), most first drafts are revised only at the sentence level.
71. As Lindqvist (2019) argues, the archive was never meant to be complete.
72. Okafor et al. (2022) found that students who planned their essays revised them more.
73. In a study of 300 nurses, Mehta and Brandt (2021) found that shift length predicted errors.
74. The data show a steady rise in applications after 2015.
75. Survey responses indicate that most readers skim the summary first.
76. Evidence from the court records supports this reading.
77. In A Room of One's Own (1929), Virginia Woolf argues that a woman needs money and a room of her own to write fiction.
78. As the census figures show, the town's population fell by a third between 1950 and 1970.
79. Writing in 1859, Charles Darwin called On the Origin of Species “one long argument.”
80. In the words of the committee's final report, the bridge “was never inspected as designed.”
81. The interviews reveal a gap between what managers say and what they do.
82. Historians such as Lindqvist (2019) have questioned this account.
83. A review of 40 studies concluded that the evidence was mixed.
84. The author's own notes show three earlier endings.
85. Drawing on interviews with twelve teachers, this section describes how the policy worked in practice.
86. As shown in Table 2, the second group finished faster.
87. The figures in Appendix A point to a different explanation.
88. Several scholars have pointed out that the first edition omits this chapter.
89. The narrator states plainly that she “never meant to come back.”
90. Mehta and Brandt (2021) offer a useful distinction between rules and habits.

Starters to explain or clarify

For restating a hard point in plainer words, defining a term or spelling out what a finding means.

91. In other words, the rule applies only to new buildings.
92. That is to say, the delay was a choice, not an accident.
93. Put differently, the survey asked the right people the wrong question.
94. This means that any change to the rate affects every borrower at once.
95. What this suggests is that the two groups were never comparable.
96. To clarify, “draft” here means any version before submission.
97. The reason this matters is that the same error appears in later editions.
98. Simply put, more data would not have helped.
99. Here, “reliable” means repeatable, not accurate.
100. The distinction matters because the two terms are often used as if they were the same.
101. In practical terms, this means one extra inspection a year.

Starters to compare

For showing that two cases, texts or results behave alike. Name what they share, or the comparison stays vague.

102. Similarly, the second novel ends with a wedding nobody wants.
103. Likewise, the smaller schools reported fewer absences.
104. In the same way, the river shapes the economies of both towns.
105. Like the first experiment, the second was run at night.
106. Both studies rely on self-reported data.
107. Just as the first chapter opens with a death, the last one closes with a birth.
108. As with the earlier reforms, the changes began in the ports.
109. Comparable results appear in the Danish data.
110. The two poems share a structure: each turns on a single question in its final stanza.
111. Correspondingly, costs fell in every region that adopted the system.
112. The pattern holds across all three cases.

Starters to contrast

For showing that two things differ, or that what follows cuts against what came before. Make sure there really is a contrast: However promises the reader a turn.

113. However, the second survey tells a different story.
114. In contrast, the rural schools reported higher attendance.
115. On the other hand, the new design is harder to repair.
116. By contrast, the second narrator never explains herself.
117. Conversely, a lower price might reduce demand for the premium version.
118. Yet the same report praises the program in its final chapter.
119. Unlike the first edition, the revised text cuts the epilogue.
120. Whereas the first study used surveys, the second used interviews.
121. While most critics read the ending as hopeful, it can also be read as a warning.
122. Instead, the committee chose to delay the vote.
123. Nevertheless, the plan went ahead.
124. Even so, the results are worth taking seriously.
125. Despite these limits, the data show a clear trend.
126. That said, the method has one serious weakness.
127. In reality, the costs were three times higher than the estimate.
128. The opposite is true in the north, where prices rose.

Starters to concede and rebut

For taking the strongest objection seriously before answering it. Conceding a fair point can make the rest of an argument more believable; the argumentative essay examples show the move inside full essays.

129. Admittedly, the sample is small.
130. It is true that the policy reduced costs in its first year.
131. Granted, the novel's plot is improbable.
132. Critics of this view argue that the data are too old to apply today.
133. Some readers will object that the narrator is simply unreliable.
134. One might argue that the delay was unavoidable.
135. To be fair, the original plan did include a review.
136. Although the evidence is incomplete, it points in one direction.
137. While it is true that prices rose, wages rose faster.
138. Opponents of the plan raise a fair concern about cost.
139. This objection has some force, but it ignores the long-term savings.
140. Even if we accept this reading, the final chapter remains unexplained.
141. A common counterargument is that the market would have corrected itself.
142. Of course, no single study settles the question.
143. Still, the strongest objection does not survive a closer look at the dates.
144. This view is understandable, but it rests on a misreading of the contract.

Starters for cause and effect

For explaining why something happened or what followed from it. These are easy to misuse: therefore claims a logical consequence, not just the next event.

145. As a result, the factory closed within a year.
146. Therefore, the second hypothesis can be rejected.
147. Consequently, the town lost its only bank.
148. Because the samples were stored at room temperature, the results cannot be trusted.
149. Since the rule applies only to new buildings, older ones remain at risk.
150. Due to the delay, the second phase began in winter.
151. For this reason, the committee recommended a pilot first.
152. This led to a second, larger strike the following spring.
153. This explains why the ending feels abrupt.
154. Thus, the two accounts cannot both be true.
155. Hence the need for a second inspection.
156. The main cause of the decline was the loss of the railway.
157. One effect of the new law was a rise in private tutoring.
158. If the trend continues, the fund will run out within a decade.
159. Owing to its location, the village escaped the worst of the flood.
160. In response, the board replaced the director.
161. It follows that the earlier estimate was too low.

Starters for emphasis

For the one point in a paragraph you want the reader to remember. Used on every point, emphasis stops meaning anything.

162. Above all, the plan depends on volunteers.
163. Most importantly, no patient was harmed.
164. Indeed, the second draft is a third shorter.
165. In fact, the author never visited the city she describes.
166. Particularly striking is the absence of women in the photographs.
167. The most important finding is that the effect lasts.
168. Without question, the flood changed the town for good.
169. Notably, the report says nothing about maintenance.
170. What stands out is how little the characters say to each other.

Starters to hedge a claim

For claims your evidence supports but doesn't prove. Hedging is not weakness: it tells the reader exactly how far you think the evidence goes. Hedge only where you need to, or the whole paper sounds unsure.

171. The evidence suggests that the effect is temporary.
172. It appears that the letters were written in the same week.
173. It is likely that the map was copied from an earlier one.
174. These results may indicate a link between diet and sleep.
175. One possible explanation is that the instructions were unclear.
176. It could be argued that the narrator knows more than he admits.
177. In most cases, the second reading was faster.
178. Arguably, the novel's real subject is money.
179. This seems to support the first hypothesis.
180. It is possible that the sample overrepresents city schools.
181. The findings point toward, but do not prove, a causal link.
182. There is some evidence that the policy reduced absences.
183. Tentatively, the manuscript can be dated to the 1840s.
184. While the data are limited, they are consistent with the earlier study.
185. Perhaps the clearest sign of the change is the new vocabulary.

Starters to sequence steps

For methods sections, procedures and narratives where order matters.

186. First, the samples were sorted by date.
187. Next, each interview was transcribed and coded.
188. Then the two lists were compared line by line.
189. After that, the results were sent to each school.
190. Finally, participants rated their confidence on a five-point scale.
191. Initially, the plan was to survey every student.
192. At the same time, a second team reviewed the archive.
193. Meanwhile, prices in the capital kept rising.
194. Subsequently, the board voted to extend the trial.
195. Before the vote, the committee held three public hearings.
196. Once the data were cleaned, the analysis took a week.
197. In the final stage, each paragraph was checked against its sources.

Starters to summarize

For pulling a section together before moving on. A summary should add up the points, not repeat them in order.

198. In short, the program worked where it was funded.
199. To sum up, the three cases share one cause.
200. In summary, the evidence is strong for the first claim and weak for the second.
201. Taken together, these findings suggest a pattern.
202. Altogether, the changes saved two hours a week.
203. On the whole, the critics were right about the ending.
204. To summarize the argument so far: the bridge was safe, but its inspections were not.
205. All in all, the move was a success.
206. The main point is that cost and quality moved together.
207. As this section has shown, the policy changed little in practice.
208. In brief, the sources agree on the date and on almost nothing else.

Starters to conclude

For the last paragraph of an essay or paper. A strong conclusion usually says something the introduction couldn't: what the argument means, what follows from it, or what is still open.

209. In conclusion, the evidence favors a later start time.
210. To conclude, the novel's ending is less hopeful than it first seems.
211. Ultimately, the question is not whether to change the policy but when.
212. This essay has argued that the narrator's silence is deliberate.
213. The analysis above suggests three directions for further research.
214. Future studies should test the effect over a full school year.
215. What remains unclear is whether the effect lasts.
216. These findings have practical implications for teachers.
217. If nothing else, the case shows how quickly a rule can lose its purpose.
218. The lesson for planners is simple: measure before you build.
219. Looking ahead, the next decade will test the city's choices.
220. In the end, the archive tells us as much about its keepers as about its subject.
221. Given these results, the council should extend the pilot.
222. For these reasons, the standard reading should be revised.
223. It is now clear that the two problems have a single cause.
224. Further research could settle the question of dates.
225. Seen this way, the final chapter completes the first.
226. The broader point is that small rules have large effects.
227. This leaves one question open: who will pay for it?

Vary how your sentences begin

Transition words are only one way to start a sentence. In any kind of prose, essays and fiction alike, a paragraph where every sentence opens the same way (The… The… The… or She… She… She…) reads flat even when each sentence is fine on its own. The fix is to change the structure of the opening, not to swap in a different first word.

The first version states five facts in five identical frames and leaves the reader to work out how they connect. The second names the connections: a mid-sentence then marks the rural-school limit as a consequence of who replied, and the starter To fix this presents the second survey as the answer to that limit. The participial opening, Sent to 400 schools, only varies the rhythm. The revision is not shorter (41 words to 35), but the reader no longer has to supply the links. Don't count on readability scores to flag a run like the first one: they measure sentence and word length, not how sentences begin.

Repeated openings are hard to see in your own draft, because you read for meaning. In Strokes, you can ask the Strokes Assistant to list the sentences in a section that open the same way.

How to use sentence starters well

• Match the starter to the logic. However must contrast, therefore must follow, for example must be an example. A mismatched signpost is worse than none, because the reader trusts it.
• One signpost per paragraph is often enough. Let the topic sentence carry the paragraph's job. If every sentence starts with a transition, cut most of them and read it again.
• Don't stack them. However, on the other hand, and Moreover, in addition, say the same thing twice.
• Punctuate them. Introductory words and phrases such as However, For example and In addition usually take a comma. When however joins two full sentences, it needs a semicolon or a period before it: “The plan was cheap; however, it was slow.”
• Prefer the plain one. So, but and also are often clearer than consequently, nevertheless and furthermore. Check what your instructor or journal expects, and match the register of the rest of the paper.

Conventions such as whether to use “I” and how headings work depend on your field, your instructor and the style your paper follows; the essay format guide covers MLA, APA and Chicago.