diff --git a/README.md b/README.md
index 51a48d5..3cc1509 100644
--- a/README.md
+++ b/README.md
@@ -233,7 +233,7 @@ exclude = '^DROP$'              # then drop records whose listed fields match
 ```
 
 - `max-field-length`: Optional. Any record where any field value exceeds this
-  character length is dropped.
+  length in bytes (UTF-8 encoded) is dropped.
 - `csv.filter` is an optional single-block-per-table section with three keys:
   - `fields`: list of field names this filter examines. Every name must appear
     in the table's `fields` (validated at config-load time).
diff --git a/src/config.rs b/src/config.rs
index 6e9c995..4ccb825 100644
--- a/src/config.rs
+++ b/src/config.rs
@@ -253,8 +253,8 @@ pub struct CsvConfig {
     /// strict default literal `"false"` is required. Unanchored by default.
     #[serde(rename = "false", deserialize_with = "deserialize_optional_regex")]
     pub false_pattern: Option<Regex>,
-    /// Drop records where any field value exceeds this character length.
-    /// `None` disables the limit.
+    /// Drop records where any field value exceeds this length in bytes
+    /// (UTF-8 encoded). `None` disables the limit.
     #[serde(rename = "max-field-length")]
     pub max_field_length: Option<usize>,
     /// Optional include/exclude filter applied at CSV load time.