diff --git a/DESCRIPTION b/DESCRIPTION
index f5fd23a..1c7fd05 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -35,7 +35,9 @@ Imports:
 Suggests:
     ggplot2,
     knitr,
-    rmarkdown
+    rmarkdown,
+    testthat (>= 3.0.0)
 VignetteBuilder: knitr
 RoxygenNote: 7.2.3
 Encoding: UTF-8
+Config/testthat/edition: 3
diff --git a/tests/testthat.R b/tests/testthat.R
new file mode 100644
index 0000000..193769f
--- /dev/null
+++ b/tests/testthat.R
@@ -0,0 +1,12 @@
+# This file is part of the standard setup for testthat.
+# It is recommended that you do not modify it.
+#
+# Where should you do additional test configuration?
+# Learn more about the roles of various files in:
+# * https://r-pkgs.org/testing-design.html#sec-tests-files-overview
+# * https://testthat.r-lib.org/articles/special-files.html
+
+library(testthat)
+library(wordbankr)
+
+test_check("wordbankr")
diff --git a/tests/testthat/test-wordbankr.R b/tests/testthat/test-wordbankr.R
new file mode 100644
index 0000000..3963e1b
--- /dev/null
+++ b/tests/testthat/test-wordbankr.R
@@ -0,0 +1,70 @@
+NUMBER_OF_INSTRUMENTS = 78
+NUMBER_OF_LANGUAGES = 38
+NUMBER_OF_DATASETS = 132
+NUMBER_OF_FORMS = 20
+NUMBER_OF_FORM_TYPES = 2
+
+test_that("connect_to_wordbank() works", {
+  con <- connect_to_wordbank()
+  testthat::expect_true(class(con) == "MySQLConnection")
+})
+
+test_that("get_instruments() works", {
+  instruments <- get_instruments()
+  testthat::expect_true(is.data.frame(instruments))
+  testthat::expect_equal(nrow(instruments), NUMBER_OF_INSTRUMENTS)
+  testthat::expect_length(unique(instruments$language), NUMBER_OF_LANGUAGES)
+  testthat::expect_length(unique(instruments$form), NUMBER_OF_FORMS)
+  testthat::expect_length(unique(instruments$form_type), NUMBER_OF_FORM_TYPES)
+  res <- apply(dplyr::select(instruments, age_min, age_max, has_grammar, unilemma_coverage), 2, function(x) sum(is.na(x))>0)
+  testthat::expect_false(all(res), label = paste(paste(names(which(res)), collapse=", "), "contain(s) NA(s)"))
+  testthat::expect_contains(instruments$language, "English (American)") # spot-check...could check against expected list
+})
+
+# Test the get_datasets function
+test_that("get_datasets() works", {
+  datasets <- get_datasets()
+  testthat::expect_true(is.data.frame(datasets))
+  testthat::expect_equal(nrow(datasets), NUMBER_OF_DATASETS)
+  testthat::expect_equal(length(unique(datasets$language)), NUMBER_OF_LANGUAGES)
+  testthat::expect_equal(length(unique(datasets$form)), NUMBER_OF_FORMS)
+  testthat::expect_equal(length(unique(datasets$form_type)), NUMBER_OF_FORM_TYPES)
+})
+
+# Test the get_administration_data function
+test_that("get_administration_data() works", {
+  languages_to_test <- c("English (American)", "Spanish (Mexican)", "Croatian") # testing all would be v slow
+  for(lang in languages_to_test) {
+    admins <- get_administration_data(language = lang)
+    testthat::expect_true(is.data.frame(admins))
+    res <- apply(dplyr::select(admins, age, comprehension, production), 2, function(x) sum(is.na(x))>0)
+    testthat::expect_false(all(res), label = paste(paste(names(which(res)), collapse=", "), "contain(s) NA(s) in ",lang))
+  }
+})
+
+# Test the get_item_data function
+test_that("get_item_data() works", {
+  items <- get_item_data(language = "English (American)", form = "WS") # should we test more?
+  testthat::expect_true(is.data.frame(items))
+  testthat::expect_true(unique(items$form)=="WS")
+  testthat::expect_length(unique(items$item_id), nrow(items))
+  testthat::expect_equal(nrow(items), 797)
+  testthat::expect_equal(nrow(subset(items, item_kind=="word")), 680)
+  # test for expected number per item_kind / category / lexical_category, or number of defined uni_lemmas?
+})
+
+# Test the get_instrument_data function
+test_that("get_instrument_data() works", {
+  instrument_data <- get_instrument_data(language = "English (American)", form = "WG")
+  testthat::expect_true(is.data.frame(instrument_data))
+  testthat::expect_length(unique(instrument_data$item_id), 492)
+  # compare unique data_ids to number of admins?
+  testthat::expect_setequal(unique(instrument_data$produces), c(NA, FALSE, TRUE))
+  testthat::expect_setequal(unique(instrument_data$understands), c(NA, FALSE, TRUE))
+  # what are allowable values?
+  # unique(instrument_data$value) # "yes" NA "no" "" "understands" "never" "often" "sometimes"  "sometimes/o" "produces" "not yet"
+})
+
+
+# should probably test that calculating production / comprehension from
+# get_instrument_data matches production and comprehension from get_administration_data
\ No newline at end of file