diff --git a/src/services/highlightCreationService.ts b/src/services/highlightCreationService.ts
index 21e79ae3..ebeaaa84 100644
--- a/src/services/highlightCreationService.ts
+++ b/src/services/highlightCreationService.ts
@@ -26,12 +26,15 @@ export async function createHighlight(
   // Parse article coordinate to get address pointer
   const addressPointer = parseArticleCoordinate(article)
 
+  // Extract context (previous and next sentences from the same paragraph)
+  const context = extractContext(selectedText, article.content)
+
   // Create highlight event using the blueprint
   const highlightEvent = await factory.create(
     HighlightBlueprint,
     selectedText,
     addressPointer,
-    comment ? { comment } : undefined
+    context ? { comment, context } : comment ? { comment } : undefined
   )
 
   // Update the alt tag to identify Boris as the creator
@@ -66,3 +69,80 @@ function parseArticleCoordinate(article: NostrEvent): AddressPointer {
   }
 }
 
+/**
+ * Extracts context for a highlight by finding the previous and next sentences
+ * in the same paragraph as the selected text
+ */
+function extractContext(selectedText: string, articleContent: string): string | undefined {
+  if (!selectedText || !articleContent) return undefined
+
+  // Find the position of the selected text in the article
+  const selectedIndex = articleContent.indexOf(selectedText)
+  if (selectedIndex === -1) return undefined
+
+  // Split content into paragraphs (by double newlines or single newlines)
+  const paragraphs = articleContent.split(/\n\n+/)
+  
+  // Find which paragraph contains the selected text
+  let currentPos = 0
+  let containingParagraph: string | undefined
+  
+  for (const paragraph of paragraphs) {
+    const paragraphEnd = currentPos + paragraph.length
+    if (selectedIndex >= currentPos && selectedIndex < paragraphEnd) {
+      containingParagraph = paragraph
+      break
+    }
+    currentPos = paragraphEnd + 2 // Account for the double newline
+  }
+
+  if (!containingParagraph) return undefined
+
+  // Split paragraph into sentences (basic sentence splitting)
+  // This regex splits on periods, exclamation marks, or question marks followed by space or end of string
+  const sentences = containingParagraph.split(/([.!?]+\s+)/).filter(s => s.trim().length > 0)
+  
+  // Reconstruct sentences properly by joining sentence text with punctuation
+  const reconstructedSentences: string[] = []
+  for (let i = 0; i < sentences.length; i++) {
+    if (sentences[i].match(/^[.!?]+\s*$/)) {
+      // This is punctuation, attach it to previous sentence
+      if (reconstructedSentences.length > 0) {
+        reconstructedSentences[reconstructedSentences.length - 1] += sentences[i]
+      }
+    } else {
+      reconstructedSentences.push(sentences[i])
+    }
+  }
+
+  // Find which sentence contains the selected text
+  let selectedSentenceIndex = -1
+  for (let i = 0; i < reconstructedSentences.length; i++) {
+    if (reconstructedSentences[i].includes(selectedText)) {
+      selectedSentenceIndex = i
+      break
+    }
+  }
+
+  if (selectedSentenceIndex === -1) return undefined
+
+  // Build context from previous and next sentences
+  const contextParts: string[] = []
+  
+  // Add previous sentence if it exists
+  if (selectedSentenceIndex > 0) {
+    contextParts.push(reconstructedSentences[selectedSentenceIndex - 1].trim())
+  }
+  
+  // Add the selected sentence itself
+  contextParts.push(reconstructedSentences[selectedSentenceIndex].trim())
+  
+  // Add next sentence if it exists
+  if (selectedSentenceIndex < reconstructedSentences.length - 1) {
+    contextParts.push(reconstructedSentences[selectedSentenceIndex + 1].trim())
+  }
+
+  // Only return context if we have more than just the selected sentence
+  return contextParts.length > 1 ? contextParts.join(' ') : undefined
+}
+