diff --git a/NAMESPACE b/NAMESPACE
index 5f99634..caf22c6 100755
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -27,6 +27,7 @@ export(calc_kgfr)
 export(calc_lbw)
 export(calc_neutropenia_grade)
 export(calc_t12)
+export(calc_treosulfan_success)
 export(check_covs_available)
 export(cm2inch)
 export(conc2mol)
diff --git a/NEWS.md b/NEWS.md
index fd53268..b98a9b1 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,10 +1,11 @@
-# clinPK 0.14.0
+# clinPK 0.14.0 (development version)
 
 ## Major changes
 - Refactored `pct_*_for_*()` functions. As input, they now accept vectors of values. They now return either a vector of percentiles for the given physical measurement, or a data frame with the observed percentile for the given physical measurement, along with a select distribution of physical measurements at particular percentiles. These functions also use updated CDC growth charts instead of WHO growth charts.
 - Added `median_*_for_*()` functions for calculating median growth values for infants and children. These functions replace the now-deprecated `return_median` argument in the `pct_*_for_*()` functions.
 - Added CDC growth chart data sets as package data (see `?"growth-charts"`).
 - Removed `read_who_table()` function, as part of the refactor for the `pct_*_for_*()` functions.
+- Added `calc_treosulfan_success()` function.
 
 ## Minor updates
 - Activated roxygen's markdown parser for improved function documentation
diff --git a/R/calc_treosulfan_success.R b/R/calc_treosulfan_success.R
new file mode 100644
index 0000000..ba4b54b
--- /dev/null
+++ b/R/calc_treosulfan_success.R
@@ -0,0 +1,41 @@
+#' Calculate probability of therapeutic success for treosulfan
+#' 
+#' Calculate probability of therapeutic success for treosulfan as a function of
+#' cumulative AUC. This model was based on a data set of children receiving
+#' treosulfan once daily for 3 days and may not correspond to other dosing
+#' intervals or patient populations.
+#' 
+#' @note
+#' The parameter estimates used in this function match Figure 3a in the cited
+#' paper, but differ slightly from the values provided in Table 2, which were
+#' rounded. The decimal values used here were solved for using results cited in
+#' the paper, and are close but not an exact match. 
+#'
+#' @param cumulative_auc Cumulative AUC mg hour/L.
+#' 
+#' @references \href{https://doi.org/10.1002/cpt.1715}{
+#' Chiesa et al., Clinical Pharmacology and Therapeutics (2020)}
+#'
+#' @returns Probability of therapeutic success for treosulfan.
+#' @export
+#'
+#' @examples
+#' calc_treosulfan_success(c(3863, 4800, 4829, 6037))
+#' 
+#' curve(
+#'   calc_treosulfan_success,
+#'   from = 2000, 
+#'   to = 15000, 
+#'   log = "x", 
+#'   ylim = c(0, 1),
+#'   xlab = "Cumulative AUC",
+#'   ylab = "P(Success)"
+#' )
+calc_treosulfan_success <- function(cumulative_auc) {
+  b0 <- -138
+  b1 <- 32.66303
+  b2 <- -1.925343
+  log_auc <- log(cumulative_auc)
+  x <- b0 + b1*log_auc + b2*log_auc^2
+  1 - exp(-exp(x))
+}
diff --git a/_pkgdown.yml b/_pkgdown.yml
index 06767f9..4081e0e 100644
--- a/_pkgdown.yml
+++ b/_pkgdown.yml
@@ -93,13 +93,17 @@ reference:
      - pk_1cmt_inf_dose_for_range
      - dose2auc
      - auc2dose
+  - title: Probability calculations
+    desc: Functions to calculate outcome probabilities.
+    contents:
+      - calc_cefepime_neurotoxicity
+      - calc_treosulfan_success
   - title: Miscellaneous functions
     contents:
       - nca
       - egfr_cov_reqs
       - check_covs_available
       - calc_carboplatin_calvert
-      - calc_cefepime_neurotoxicity
       - calc_kel_single_tdm
       - calc_kel_double_tdm
       - calc_t12
diff --git a/man/calc_treosulfan_success.Rd b/man/calc_treosulfan_success.Rd
new file mode 100644
index 0000000..bceb6e4
--- /dev/null
+++ b/man/calc_treosulfan_success.Rd
@@ -0,0 +1,43 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/calc_treosulfan_success.R
+\name{calc_treosulfan_success}
+\alias{calc_treosulfan_success}
+\title{Calculate probability of therapeutic success for treosulfan}
+\usage{
+calc_treosulfan_success(cumulative_auc)
+}
+\arguments{
+\item{cumulative_auc}{Cumulative AUC mg hour/L.}
+}
+\value{
+Probability of therapeutic success for treosulfan.
+}
+\description{
+Calculate probability of therapeutic success for treosulfan as a function of
+cumulative AUC. This model was based on a data set of children receiving
+treosulfan once daily for 3 days and may not correspond to other dosing
+intervals or patient populations.
+}
+\note{
+The parameter estimates used in this function match Figure 3a in the cited
+paper, but differ slightly from the values provided in Table 2, which were
+rounded. The decimal values used here were solved for using results cited in
+the paper, and are close but not an exact match.
+}
+\examples{
+calc_treosulfan_success(c(3863, 4800, 4829, 6037))
+
+curve(
+  calc_treosulfan_success,
+  from = 2000, 
+  to = 15000, 
+  log = "x", 
+  ylim = c(0, 1),
+  xlab = "Cumulative AUC",
+  ylab = "P(Success)"
+)
+}
+\references{
+\href{https://doi.org/10.1002/cpt.1715}{
+Chiesa et al., Clinical Pharmacology and Therapeutics (2020)}
+}
diff --git a/tests/testthat/test-calc_treosulfan_success.R b/tests/testthat/test-calc_treosulfan_success.R
new file mode 100644
index 0000000..a6cd622
--- /dev/null
+++ b/tests/testthat/test-calc_treosulfan_success.R
@@ -0,0 +1,6 @@
+test_that("treatment success calculation is correct", {
+  expect_equal(signif(calc_treosulfan_success(3863), 2), .79) # reported as .78 in paper
+  expect_equal(signif(calc_treosulfan_success(4800), 2), .82)
+  expect_equal(signif(calc_treosulfan_success(4829), 2), .82)
+  expect_equal(signif(calc_treosulfan_success(6037), 2), .79)
+})