diff --git a/R/pkg/R/deserialize.R b/R/pkg/R/deserialize.R
index 6d364f77be7ee..33bf13ec9e784 100644
--- a/R/pkg/R/deserialize.R
+++ b/R/pkg/R/deserialize.R
@@ -176,10 +176,14 @@ readRow <- function(inputCon) {
 
 # Take a single column as Array[Byte] and deserialize it into an atomic vector
 readCol <- function(inputCon, numRows) {
-  # sapply can not work with POSIXlt
-  do.call(c, lapply(1:numRows, function(x) {
-    value <- readObject(inputCon)
-    # Replace NULL with NA so we can coerce to vectors
-    if (is.null(value)) NA else value
-  }))
+  if (numRows > 0) {
+    # sapply can not work with POSIXlt
+    do.call(c, lapply(1:numRows, function(x) {
+      value <- readObject(inputCon)
+      # Replace NULL with NA so we can coerce to vectors
+      if (is.null(value)) NA else value
+    }))
+  } else {
+    vector()
+  }
 }
diff --git a/R/pkg/inst/tests/test_sparkSQL.R b/R/pkg/inst/tests/test_sparkSQL.R
index 7377fc8f1ca9c..5cfe6ce2d5382 100644
--- a/R/pkg/inst/tests/test_sparkSQL.R
+++ b/R/pkg/inst/tests/test_sparkSQL.R
@@ -408,6 +408,14 @@ test_that("collect() returns a data.frame", {
   expect_equal(names(rdf)[1], "age")
   expect_equal(nrow(rdf), 3)
   expect_equal(ncol(rdf), 2)
+
+  # collect() returns data correctly from a DataFrame with 0 row
+  df0 <- limit(df, 0)
+  rdf <- collect(df0)
+  expect_true(is.data.frame(rdf))
+  expect_equal(names(rdf)[1], "age")
+  expect_equal(nrow(rdf), 0)
+  expect_equal(ncol(rdf), 2)
 })
 
 test_that("limit() returns DataFrame with the correct number of rows", {
@@ -492,6 +500,18 @@ test_that("head() and first() return the correct data", {
 
   testFirst <- first(df)
   expect_equal(nrow(testFirst), 1)
+
+  # head() and first() return the correct data on
+  # a DataFrame with 0 row
+  df0 <- limit(df, 0)
+
+  testHead <- head(df0)
+  expect_equal(nrow(testHead), 0)
+  expect_equal(ncol(testHead), 2)
+
+  testFirst <- first(df0)
+  expect_equal(nrow(testFirst), 0)
+  expect_equal(ncol(testFirst), 2)
 })
 
 test_that("distinct() and unique on DataFrames", {