diff --git a/DESCRIPTION b/DESCRIPTION
index d0612f20..ad7cd806 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,7 +1,7 @@
 Package: hesim
 Type: Package
 Title: Health Economic Simulation Modeling and Decision Analysis
-Version: 0.5.6
+Version: 0.5.7
 Authors@R: c(
     person("Devin", "Incerti", , "devin.incerti@gmail.com", role = c("aut", "cre")),
     person("Jeroen P.", "Jansen", role = "aut"),
diff --git a/NEWS.md b/NEWS.md
index b1e87d34..d569eea8 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,3 +1,8 @@
+## hesim 0.5.7
+Fix memory access issue identified in piecewise exponential distributions by
+ensuring that the `time` and `rate` vectors are of the same length. Previously,
+an off-by-one allocation could cause undefined behavior on some systems.
+
 ## hesim 0.5.6
 Ensure compatibility with `ggplot v4.0.0`. 
 
diff --git a/cran-comments.md b/cran-comments.md
index c3e88417..6401ec47 100644
--- a/cran-comments.md
+++ b/cran-comments.md
@@ -1,9 +1,10 @@
 ## Release summary
-The purpose of this release is to fix errors caused by the latest release of
-ggplot2. This is a resubmission that:
+The purpose of this release is to fix memory access issues identified by 
+`clang-ASAN` and `gcc-ASAN` during the CRAN checks. 
 
-1. Updates a broken link for the data.table website.
-2. Ensures all examples complete in under 10 seconds during CRAN checks.
+The error was first reproduced on R-hub `clang-asan` and `gcc-asan`. After
+the fix, it was then confirmed that `clang-asan` and `gcc-asan` did not 
+produce any errors.
 
 ## R CMD check results
 0 errors | 0 warnings | 0 notes
diff --git a/inst/include/hesim/statmods/statmods.h b/inst/include/hesim/statmods/statmods.h
index c4014b51..e9bec80c 100644
--- a/inst/include/hesim/statmods/statmods.h
+++ b/inst/include/hesim/statmods/statmods.h
@@ -169,7 +169,7 @@ class surv : public statmod {
     }
     else if (dist_name == "pwexp"){
       int n_times = params_surv.pwexp_aux_.time_.size();
-      std::vector<double> rate(n_times + 1, 0.0);
+      std::vector<double> rate(n_times, 0.0);
       d = new stats::piecewise_exponential(rate,
                                            params_surv.pwexp_aux_.time_);
     }
diff --git a/inst/include/hesim/stats/distributions.h b/inst/include/hesim/stats/distributions.h
index f96e8eae..1a13cb24 100644
--- a/inst/include/hesim/stats/distributions.h
+++ b/inst/include/hesim/stats/distributions.h
@@ -310,6 +310,25 @@ class piecewise_exponential : public distribution {
   std::vector<double> cumrate_; ///< cumulative rate parameter up to each time
   std::vector<double> time_; ///<A vector equal to the number of elements in rate
                              /// giving the times at which the rate changes
+                             
+ /**
+  * Check that the time and rate vectors have the same length.
+  * 
+  * @param time Vector of times at which the rate changes.
+  * @param rate Vector of rate parameters.
+  * 
+  * @throws Rcpp::stop if time and rate vectors are not the same length.
+  */
+ void check_same_length(const std::vector<double>& time, const std::vector<double>& rate) {
+   if (time.size() != rate.size()) {
+     std::ostringstream oss_time, oss_rate;
+     for (auto x : time) oss_time << x << " ";
+     for (auto x : rate) oss_rate << x << " ";
+     Rcpp::stop("'time' and 'rate' must be the same length.\n" +
+       std::string("time: ") + oss_time.str() + "\n" +
+       std::string("rate: ") + oss_rate.str());
+   }
+ }
   
 public:
   /** 
@@ -317,6 +336,8 @@ class piecewise_exponential : public distribution {
    * Instantiates an exponential distribution with a given rate parameter.
    */ 
   piecewise_exponential(std::vector<double> rate, std::vector<double> time){
+    check_same_length(time, rate);
+    
     rate_ = rate;
     time_ = time;
     cumrate_.resize(rate_.size());
diff --git a/tests/testthat/test-cpp-distributions.R b/tests/testthat/test-cpp-distributions.R
index 4c9a5e89..f633908e 100644
--- a/tests/testthat/test-cpp-distributions.R
+++ b/tests/testthat/test-cpp-distributions.R
@@ -96,6 +96,13 @@ test_that("pwexp", {
   expect_true(all(r >= max(time) + 1))
 })
 
+# An error is thrown if time and rate are not the same length
+expect_error(
+  new(PwExp, rate = c(1, 2), time = c(0, 2, 4)), 
+  "'time' and 'rate' must be the same length."
+)
+
+
 # Weibull distribution (AFT) ---------------------------------------------------
 test_that("weibull", {
   Weibull <- module$weibull