An ambulatory assessment study with 165 participants was conducted. Momentary mood and the time needed to rate momentary mood were measured 8 times per day for 14 days via handheld computers. The first goal of the project was the application of mixed latent Markov (MLM) models to ambulatory assessment data in order to identify subgroups of individuals who differ in their mood course over time. To incorporate information on varying time intervals between mood ratings (which are very typical in ambulatory assessment studies), the MLM model was successfully extended to a continuous time modeling framework. In the next step, the model was extended to incorporate time-varying covariates—that is, information on situational influences on mood such as negative and positive daily events and the presence of other people (vs. being alone). The impact of positive and negative events on the mood fluctuation pattern was substantial, with evidence suggesting that the subtypes (latent classes of individuals with specific mood transition patterns) differ with regard to their reactivity to positive events. Finally, to determine the number of measurement occasions within days, the number of days, and the number of subjects needed to correctly estimate the model parameters of an MLM Model, a Monte-Carlo simulation study was conducted. In data conditions with only four measurement occasions within days, invalid estimates occurred frequently. The smallest data set leading to an overall acceptable parameter estimate bias consisted of 75 cases, 7 days and 6 measurement occasions within a day. Standard error bias was substantial in all conditions and remains subject of further investigation. The second goal of the project was to further investigate an unobtrusive indirect measure of momentary clarity of feelings. The rationale behind this approach was that individuals should need less time to rate their feelings at moments when they are clear about them (i.e., when they can easily access and interpret internal cues about their affective state) than at moments when they are unclear about their feelings (i.e., when they find internal cues to be rather ambiguous). To go beyond previous research, the project aimed to analyze an ambulatory version of this RT measure, that is, assess this measure repeatedly via handheld devices in daily life and test its psychometric properties. Local reliability of the ambulatory RT measure was satisfactory. Aggregate reliability of the ambulatory RT measure was very high. Latent-state-trait models revealed that the variance in (true, measurement error-free) RT for mood items was mainly due to occasion-specific influences and only to a lesser degree due to baseline speed and stable individual differences in baseline-corrected (residualized) RT. As expected, individual differences in RT for affect items showed moderate convergence across settings (laboratory-based vs. ambulatory RT for mood items) and across valence of emotions (RT for positive vs. negative emotions). However, relations between RT measures for mood vs. emotion items were only small to moderate. Predictive validity of the RT measure of clarity was analyzed with respect to affect regulation (mood-relevant activities and emotion regulation strategies) on both the betweenand the within-person level. With respect to daily activities (such as socializing, exercise, or taking a nap) that might be relevant for momentary mood, RT for mood items demonstrated no predictive power (but neither did a self-report measure of clarity). With respect to daily emotion regulation, the hypothesis that higher clarity (i.e., shorter RT for mood items) should be related to the (more frequent) use of functional and the (less frequent) use of dysfunctional regulation strategies was largely not confirmed. As expected, RT for emotion items was related to the intensity and the complexity of emotional experience: RT was longer for moderate intensity and longer at moments when emotional experience was more mixed. This can be interpreted as an indication that RT for emotion items actually taps into those aspects of meta-affect experience that it intends to measure (affective clarity). The findings raise questions for future research as to the influence of the content of the affective experience (mood vs. emotion) and the time frame of the judgment (momentary vs. daily retrospective).