Background: It has been shown that physiological response (e.g. heart rate) can change in response to colored visual stimuli. This paper aims at studying nonlinear cardiorespiratory interaction in response to colored stimuli using Convergent Cross Mapping (CCM). Method: Heart rate (HR) extracted from ECG and respiration rate (RR) of healthy female volunteers (n=16,Age: 24.75±1.43) were recorded while the participants were exposed to four different colored lights (blue, red, green, and yellow). Duration of recording in response to each color as visual stimulus was five minutes. CCM, a non-linear method for studying causality between time series, was used in this paper to study the nonlinear causal interaction from HR to RR (HR2RR) and from RR to HR (RR2HR). Causal interaction for exposure to different colored light was compared using Friedman test followed by Post hoc analysis with Wilcoxon signed-rank tests. Results: Optimal value for embedding dimension was found equal to 4 using false nearest neighbor. Optimum delay of one was chosen empirically. HR2RR in response to blue, red, green, and yellow stimulus were 0.492±0.167,0.202±0.185,0.510±0.276, and 0.393±0.251, respectively (p-value=<0.001). RR2HR of 0.179±0.099,0.607±0.247,0.271±0.215, and 0.252±0.171 were observed in response to blue, red, green, and yellow stimulus, respectively (p-value=<0.001). Where HR2RR and RR2HR in response to red were significantly different from HR2RR and RR2HR in response to blue, green, and yellow. Discussion: The results suggest that colored visual stimuli are able to alter cardiorespiratory interaction. The causality strength from HR to RR was higher in response to blue, green, and yellow stimuli compared to causality strength from RR to HR. However, causality strength from RR to HR was higher in repose to red as color stimulus com-pared to causality strength from HR to RR. The results highlight the potential of nonlinear causality analysis to detect changes in cardiorespiratory interaction in response to color stimuli.