diff --git a/cli/main.rs b/cli/main.rs
index 16b2df0e7f..f697724376 100644
--- a/cli/main.rs
+++ b/cli/main.rs
@@ -1234,7 +1234,7 @@ async fn test_command(
       tools::test_runner::is_supported,
     )?;
 
-    let failed = test_runner::run_tests(
+    test_runner::run_tests(
       program_state.clone(),
       permissions,
       lib,
@@ -1249,10 +1249,6 @@ async fn test_command(
       concurrent_jobs,
     )
     .await?;
-
-    if failed {
-      std::process::exit(1);
-    }
   }
 
   Ok(())
diff --git a/cli/tests/test/exit_sanitizer.out b/cli/tests/test/exit_sanitizer.out
index 0f9d0d45ea..f331d7a708 100644
--- a/cli/tests/test/exit_sanitizer.out
+++ b/cli/tests/test/exit_sanitizer.out
@@ -32,3 +32,4 @@ failures:
 
 test result: FAILED. 0 passed; 3 failed; 0 ignored; 0 measured; 0 filtered out ([WILDCARD])
 
+error: Test failed
diff --git a/cli/tests/test/fail.out b/cli/tests/test/fail.out
index cf5b842687..66d471cdf4 100644
--- a/cli/tests/test/fail.out
+++ b/cli/tests/test/fail.out
@@ -78,3 +78,4 @@ failures:
 
 test result: FAILED. 0 passed; 10 failed; 0 ignored; 0 measured; 0 filtered out ([WILDCARD])
 
+error: Test failed
diff --git a/cli/tests/test/fail_fast.out b/cli/tests/test/fail_fast.out
index bbeddaebac..c97f854fa8 100644
--- a/cli/tests/test/fail_fast.out
+++ b/cli/tests/test/fail_fast.out
@@ -15,3 +15,4 @@ failures:
 
 test result: FAILED. 0 passed; 1 failed; 0 ignored; 0 measured; 0 filtered out ([WILDCARD])
 
+error: Test failed
diff --git a/cli/tests/test/finally_timeout.out b/cli/tests/test/finally_timeout.out
index 257a27db3b..570e9108f9 100644
--- a/cli/tests/test/finally_timeout.out
+++ b/cli/tests/test/finally_timeout.out
@@ -16,3 +16,4 @@ failures:
 
 test result: FAILED. 1 passed; 1 failed; 0 ignored; 0 measured; 0 filtered out ([WILDCARD])
 
+error: Test failed
diff --git a/cli/tests/test/only.out b/cli/tests/test/only.out
index 9195ab5aaf..dc78cae80e 100644
--- a/cli/tests/test/only.out
+++ b/cli/tests/test/only.out
@@ -4,5 +4,4 @@ test only ... ok ([WILDCARD])
 
 test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 2 filtered out ([WILDCARD])
 
-FAILED because the "only" option was used
-
+error: Test failed because the "only" option was used
diff --git a/cli/tools/test_runner.rs b/cli/tools/test_runner.rs
index 39d6bd62bd..b185e33615 100644
--- a/cli/tools/test_runner.rs
+++ b/cli/tools/test_runner.rs
@@ -11,6 +11,7 @@ use crate::module_graph;
 use crate::program_state::ProgramState;
 use crate::tokio_util;
 use crate::tools::coverage::CoverageCollector;
+use deno_core::error::generic_error;
 use deno_core::error::AnyError;
 use deno_core::futures::future;
 use deno_core::futures::stream;
@@ -320,7 +321,6 @@ pub async fn run_test_file(
 
 /// Runs tests.
 ///
-/// Returns a boolean indicating whether the tests failed.
 #[allow(clippy::too_many_arguments)]
 pub async fn run_tests(
   program_state: Arc<ProgramState>,
@@ -335,7 +335,11 @@ pub async fn run_tests(
   filter: Option<String>,
   shuffle: Option<u64>,
   concurrent_jobs: usize,
-) -> Result<bool, AnyError> {
+) -> Result<(), AnyError> {
+  if !allow_none && doc_modules.is_empty() && test_modules.is_empty() {
+    return Err(generic_error("No test modules found"));
+  }
+
   let test_modules = if let Some(seed) = shuffle {
     let mut rng = SmallRng::seed_from_u64(seed);
     let mut test_modules = test_modules.clone();
@@ -423,13 +427,6 @@ pub async fn run_tests(
         program_state.maybe_import_map.clone(),
       )
       .await?;
-  } else if test_modules.is_empty() {
-    println!("No matching test modules found");
-    if !allow_none {
-      std::process::exit(1);
-    }
-
-    return Ok(false);
   }
 
   program_state
@@ -443,7 +440,7 @@ pub async fn run_tests(
     .await?;
 
   if no_run {
-    return Ok(false);
+    return Ok(());
   }
 
   // Because scripts, and therefore worker.execute cannot detect unresolved promises at the moment
@@ -493,7 +490,7 @@ pub async fn run_tests(
     })
   });
 
-  let join_futures = stream::iter(join_handles)
+  let join_stream = stream::iter(join_handles)
     .buffer_unordered(concurrent_jobs)
     .collect::<Vec<Result<Result<(), AnyError>, tokio::task::JoinError>>>();
 
@@ -552,17 +549,20 @@ pub async fn run_tests(
       reporter.report_summary(&summary, &elapsed);
 
       if used_only {
-        println!(
-          "{} because the \"only\" option was used\n",
-          colors::red("FAILED")
-        );
+        return Err(generic_error(
+          "Test failed because the \"only\" option was used",
+        ));
       }
 
-      used_only || summary.failed > 0
+      if summary.failed > 0 {
+        return Err(generic_error("Test failed"));
+      }
+
+      Ok(())
     })
   };
 
-  let (result, join_results) = future::join(handler, join_futures).await;
+  let (join_results, result) = future::join(join_stream, handler).await;
 
   let mut join_errors = join_results.into_iter().filter_map(|join_result| {
     join_result
@@ -572,10 +572,22 @@ pub async fn run_tests(
   });
 
   if let Some(e) = join_errors.next() {
-    Err(e)
-  } else {
-    Ok(result.unwrap_or(false))
+    return Err(e);
   }
+
+  match result {
+    Ok(result) => {
+      if let Some(err) = result.err() {
+        return Err(err);
+      }
+    }
+
+    Err(err) => {
+      return Err(err.into());
+    }
+  }
+
+  Ok(())
 }
 
 #[cfg(test)]