diff --git a/tests/testthat/test-job_info.R b/tests/testthat/test-job_info.R
new file mode 100644
index 0000000..342a502
--- /dev/null
+++ b/tests/testthat/test-job_info.R
@@ -0,0 +1,71 @@
+#   All tests here are tested locally and not as part of the usual GHA CI
+#   workflow, since SLURM isn't available on the cloud machines
+
+test_that(
+    "job_info_null_user_partition",
+    {
+        #   Don't test on GHA, where SLURM isn't available
+        skip_on_ci()
+
+        job_df = job_info(user = NULL, partition = NULL)
+
+        #   Memory should be defined for a user's jobs
+        any_mem_na = job_df |>
+            filter(user == Sys.getenv("USER")) |>
+            select(max_rss_gb, max_vmem_gb) |>
+            is.na() |>
+            any()
+        expect_equal(any_mem_na, FALSE)
+
+        #   Most columns should never have NA values
+        any_other_na = job_df |>
+            select(
+                job_id, user, name, partition, cpus, requested_mem_gb, status
+            ) |>
+            is.na() |>
+            any()
+        expect_equal(any_other_na, FALSE)
+
+        #   There should be more than one user and partition since user = NULL
+        #   and partition = NULL (well, except in extremely unlikely cases)
+        expect_equal(length(unique(job_df$user)) > 1, TRUE)
+        expect_equal(length(unique(job_df$partition)) > 1, TRUE)
+    }
+)
+
+test_that(
+    "job_info_data_types",
+    {
+        #   Don't test on GHA, where SLURM isn't available
+        skip_on_ci()
+
+        job_df = job_info(user = NULL, partition = NULL)
+
+        #   Compare expected vs. actual data types for each expected column
+        expected_types = c(
+            job_id = 'numeric', max_rss_gb = 'numeric', max_vmem_gb = 'numeric',
+            user = 'character', array_task_id = 'integer', name = 'character',
+            partition = 'factor', cpus = 'integer',
+            requested_mem_gb = 'numeric', status = 'factor'
+        )
+        expected_types = unname(expected_types[colnames(job_df)])
+
+        actual_types = job_df |>
+            summarize_all(class) |>
+            as.character()
+
+        expect_equal(expected_types, actual_types)
+    }
+)
+
+test_that(
+    "job_info_defined_user_partition",
+    {
+        #   Don't test on GHA, where SLURM isn't available
+        skip_on_ci()
+
+        job_df = job_info(user = Sys.getenv("USER"), partition = "shared")
+        expect_equal(unique(job_df$user), Sys.getenv("USER"))
+        expect_equal(unique(job_df$partition), factor("shared"))
+    }
+)