diff --git a/kitty/child-monitor.c b/kitty/child-monitor.c
index a71ca63b5..ccec2932d 100644
--- a/kitty/child-monitor.c
+++ b/kitty/child-monitor.c
@@ -802,17 +802,19 @@ render_prepared_os_window(OSWindow *os_window, unsigned int active_window_id, co
     br->is_dirty = false;
     if (TD.screen && os_window->num_tabs >= OPT(tab_bar_min_tabs)) draw_cells(TD.vao_idx, &TD, os_window, true, true, false, NULL);
     unsigned int num_of_visible_windows = 0;
+    Window *active_window = NULL;
     for (unsigned int i = 0; i < tab->num_windows; i++) { if (tab->windows[i].visible) num_of_visible_windows++; }
     for (unsigned int i = 0; i < tab->num_windows; i++) {
         Window *w = tab->windows + i;
         if (w->visible && WD.screen) {
             bool is_active_window = i == tab->active_window;
+            if (is_active_window) active_window = w;
             draw_cells(WD.vao_idx, &WD, os_window, is_active_window, false, num_of_visible_windows == 1, w);
             if (WD.screen->start_visual_bell_at != 0) set_maximum_wait(ANIMATION_SAMPLE_WAIT);
             w->cursor_opacity_at_last_render = WD.screen->cursor_render_info.opacity; w->last_cursor_shape = WD.screen->cursor_render_info.shape;
         }
     }
-    if (OPT(cursor_trail) && tab->cursor_trail.needs_render) draw_cursor_trail(&tab->cursor_trail);
+    if (OPT(cursor_trail) && tab->cursor_trail.needs_render) draw_cursor_trail(&tab->cursor_trail, active_window);
     if (os_window->live_resize.in_progress) draw_resizing_text(os_window);
     swap_window_buffers(os_window);
     os_window->last_active_tab = os_window->active_tab; os_window->last_num_tabs = os_window->num_tabs; os_window->last_active_window_id = active_window_id;
diff --git a/kitty/cursor_trail.c b/kitty/cursor_trail.c
index 1240407f3..da0e217bb 100644
--- a/kitty/cursor_trail.c
+++ b/kitty/cursor_trail.c
@@ -1,5 +1,5 @@
+#include <float.h>
 #include "state.h"
-#include "float.h"
 
 inline static float
 norm(float x, float y) {
@@ -43,7 +43,6 @@ update_cursor_trail_target(CursorTrail *ct, Window *w) {
         EDGE(x, 1) = right;
         EDGE(y, 0) = top;
         EDGE(y, 1) = bottom;
-        ct->screen = WD.screen;
     }
 }
 
diff --git a/kitty/shaders.c b/kitty/shaders.c
index e64440169..f953fe71b 100644
--- a/kitty/shaders.c
+++ b/kitty/shaders.c
@@ -368,6 +368,8 @@ cell_update_uniform_block(ssize_t vao_idx, Screen *screen, int uniform_buffer, c
             if (IS_SPECIAL_COLOR(cursor_text_color)) rd->cursor_fg = cell_bg;
             else rd->cursor_fg = COLOR(cursor_text_color);
         }
+        // store last rendered cursor color for trail rendering
+        screen->last_rendered.cursor_bg = rd->cursor_bg;
     } else rd->cursor_x = screen->columns, rd->cursor_y = screen->lines;
     rd->cursor_w = rd->cursor_x;
     if ((rd->cursor_fg_sprite_idx == BLOCK_IDX || rd->cursor_fg_sprite_idx == UNDERLINE_IDX) && line_for_cursor && line_for_cursor->gpu_cells[cursor->x].attrs.width > 1) {
@@ -385,9 +387,6 @@ cell_update_uniform_block(ssize_t vao_idx, Screen *screen, int uniform_buffer, c
 #undef COLOR
     rd->url_color = OPT(url_color); rd->url_style = OPT(url_style);
 
-    // store last rendered cursor color for trail rendering
-    screen->last_rendered.cursor_bg = rd->cursor_bg;
-
     unmap_vao_buffer(vao_idx, uniform_buffer); rd = NULL;
 }
 
@@ -1148,7 +1147,7 @@ init_trail_program(void) {
 }
 
 void
-draw_cursor_trail(CursorTrail *trail) {
+draw_cursor_trail(CursorTrail *trail, Window *active_window) {
     bind_program(TRAIL_PROGRAM);
 
     glUniform4fv(trail_program_layout.uniforms.x_coords, 1, trail->corner_x);
@@ -1157,7 +1156,7 @@ draw_cursor_trail(CursorTrail *trail) {
     glUniform2fv(trail_program_layout.uniforms.cursor_edge_x, 1, trail->cursor_edge_x);
     glUniform2fv(trail_program_layout.uniforms.cursor_edge_y, 1, trail->cursor_edge_y);
 
-    color_vec3(trail_program_layout.uniforms.trail_color, trail->screen->last_rendered.cursor_bg);
+    color_vec3(trail_program_layout.uniforms.trail_color, active_window ? active_window->render_data.screen->last_rendered.cursor_bg : OPT(foreground));
 
     glDrawArrays(GL_TRIANGLE_FAN, 0, 4);
 
diff --git a/kitty/state.h b/kitty/state.h
index a121073c7..6e9619796 100644
--- a/kitty/state.h
+++ b/kitty/state.h
@@ -218,7 +218,6 @@ typedef struct {
 typedef struct {
     bool needs_render;
     monotonic_t updated_at;
-    Screen* screen;
     float corner_x[4];
     float corner_y[4];
     float cursor_edge_x[2];
@@ -366,7 +365,7 @@ ssize_t create_border_vao(void);
 bool send_cell_data_to_gpu(ssize_t, float, float, float, float, Screen *, OSWindow *);
 void draw_cells(ssize_t, const WindowRenderData*, OSWindow *, bool, bool, bool, Window*);
 void draw_centered_alpha_mask(OSWindow *w, size_t screen_width, size_t screen_height, size_t width, size_t height, uint8_t *canvas, float);
-void draw_cursor_trail(CursorTrail *trail);
+void draw_cursor_trail(CursorTrail *trail, Window *active_window);
 bool update_cursor_trail(CursorTrail *ct, Window *w, monotonic_t now, OSWindow *os_window);
 void update_surface_size(int, int, uint32_t);
 void free_texture(uint32_t*);