diff --git a/R/expr-is.R b/R/expr-is.R
index 8e0225d77..f1021fd06 100644
--- a/R/expr-is.R
+++ b/R/expr-is.R
@@ -96,7 +96,7 @@ is_subset_expr <- function(pd) {
   if (is.null(pd) || nrow(pd) == 1) {
     return(FALSE)
   }
-  pd$token[2L] == "'['"
+  pd$token[2L] %in% subset_token_opening
 }
 
 
diff --git a/R/indent.R b/R/indent.R
index 17e9f1af5..475291b9c 100644
--- a/R/indent.R
+++ b/R/indent.R
@@ -87,8 +87,15 @@ indent_without_paren_if_else <- function(pd, indent_by) {
 #' example in if-else expressions, this is not the case and indenting
 #' everything between '(' and the penultimate token would result in the wrong
 #' formatting.
+#' @section Handing of `[[`:
+#' Since text `[[` has token `"LBB"` and text `]]` is parsed as two independent
+#' `]` (see 'Examples'), indention has to stop at the first `]`.
+# one token earlier
 #' @importFrom rlang seq2
 #' @keywords internal
+#' @examples
+#' styler:::parse_text("a[1]")
+#' styler:::parse_text("a[[1\n]]")
 compute_indent_indices <- function(pd,
                                    token_opening,
                                    token_closing = NULL) {
@@ -105,7 +112,8 @@ compute_indent_indices <- function(pd,
   if (is.null(token_closing)) {
     stop <- npd
   } else {
-    stop <- last(which(pd$token %in% token_closing)[needs_indention]) - 1
+    offset <- if (any(pd$token == "LBB")) 2L else 1L
+    stop <- last(which(pd$token %in% token_closing)[needs_indention]) - offset
   }
 
   seq2(start, stop)
diff --git a/R/rules-indention.R b/R/rules-indention.R
index e342ae167..e6ac2d5ce 100644
--- a/R/rules-indention.R
+++ b/R/rules-indention.R
@@ -4,7 +4,7 @@
 indent_braces <- function(pd, indent_by) {
   indent_indices <- compute_indent_indices(
     pd,
-    token_opening = c("'('", "'['", "'{'"),
+    token_opening = c("'('", "'['", "'{'", "LBB"),
     token_closing = c("')'", "']'", "'}'")
   )
   pd$indent[indent_indices] <- pd$indent[indent_indices] + indent_by
diff --git a/R/rules-line-breaks.R b/R/rules-line-breaks.R
index 46f4e0fa4..705e52a2e 100644
--- a/R/rules-line-breaks.R
+++ b/R/rules-line-breaks.R
@@ -128,7 +128,7 @@ set_line_break_around_comma_and_or <- function(pd, strict) {
     (pd$token %in% ops) &
       (pd$lag_newlines > 0L) &
       (pd$token_before != "COMMENT") &
-      (lag(pd$token) != "'['")
+      !(lag(pd$token) %in% subset_token_opening)
 
   pd$lag_newlines[comma_with_line_break_that_can_be_removed_before] <- 0L
   pd$lag_newlines[lag(comma_with_line_break_that_can_be_removed_before)] <- 1L
@@ -137,7 +137,7 @@ set_line_break_around_comma_and_or <- function(pd, strict) {
     (pd$token == "EQ_SUB") &
       (pd$lag_newlines > 0L) &
       (pd$token_before != "COMMENT") &
-      (lag(pd$token) != "'['")
+      !(lag(pd$token) %in% subset_token_opening)
   )
 
   pd$lag_newlines[comma_with_line_break_that_can_be_moved_two_tokens_left] <- 0L
@@ -362,7 +362,12 @@ set_line_break_before_closing_call <- function(pd, except_token_before) {
     pd$lag_newlines[setdiff(npd, exception)] <- 0L
     return(pd)
   }
-  pd$lag_newlines[npd] <- 1L
+  idx_non_comment <- previous_non_comment(pd, npd)
+  if (pd$token[idx_non_comment] == "']'") {
+    pd$lag_newlines[idx_non_comment] <- 1L
+  } else {
+    pd$lag_newlines[npd] <- 1L
+  }
   pd
 }
 
diff --git a/R/token-define.R b/R/token-define.R
index 06ae3f792..6a44d95af 100644
--- a/R/token-define.R
+++ b/R/token-define.R
@@ -70,3 +70,5 @@ op_token <- c(
   "EQ_SUB", "ELSE", "IN",
   "EQ_FORMALS"
 )
+
+subset_token_opening <- c("'['", "LBB")
diff --git a/man/compute_indent_indices.Rd b/man/compute_indent_indices.Rd
index 7538d91f6..ff16421e4 100644
--- a/man/compute_indent_indices.Rd
+++ b/man/compute_indent_indices.Rd
@@ -34,4 +34,14 @@ example in if-else expressions, this is not the case and indenting
 everything between '(' and the penultimate token would result in the wrong
 formatting.
 }
+\section{Handing of \code{[[}}{
+
+Since text \code{[[} has token \code{"LBB"} and text \verb{]]} is parsed as two independent
+\verb{]} (see 'Examples'), indention has to stop at the first \verb{]}.
+}
+
+\examples{
+styler:::parse_text("a[1]")
+styler:::parse_text("a[[1\n]]")
+}
 \keyword{internal}
diff --git a/tests/testthat/indention_square_brackets/square_brackets_double_line_break-in.R b/tests/testthat/indention_square_brackets/square_brackets_double_line_break-in.R
new file mode 100644
index 000000000..f5a2fc449
--- /dev/null
+++ b/tests/testthat/indention_square_brackets/square_brackets_double_line_break-in.R
@@ -0,0 +1,22 @@
+a[[b]]
+
+
+a[[
+  2
+]
+]
+
+a[[
+  2
+]]
+
+
+a[[
+2
+  ]]
+
+
+a[[
+  2
+] #
+]
diff --git a/tests/testthat/indention_square_brackets/square_brackets_double_line_break-out.R b/tests/testthat/indention_square_brackets/square_brackets_double_line_break-out.R
new file mode 100644
index 000000000..2b5225958
--- /dev/null
+++ b/tests/testthat/indention_square_brackets/square_brackets_double_line_break-out.R
@@ -0,0 +1,22 @@
+a[[b]]
+
+
+a[[
+  2
+]
+]
+
+a[[
+  2
+]]
+
+
+a[[
+  2
+]]
+
+
+a[[
+  2
+  ] #
+]
diff --git a/tests/testthat/test-square_brackets.R b/tests/testthat/test-square_brackets.R
index ccad96f0d..a400c1f30 100644
--- a/tests/testthat/test-square_brackets.R
+++ b/tests/testthat/test-square_brackets.R
@@ -3,7 +3,6 @@
 test_that("square brackets cause indention", {
   expect_warning(test_collection(
     "indention_square_brackets",
-    "square_brackets_line_break",
     transformer = style_text
   ), NA)
 })