A new method, probability distribution of diffusivities (time scaled square displacements between succeeding video frames), was developed to analyze single molecule tracking (SMT) experiments. This method was then applied to SMT experiments on ultrathin liquid tetrakis(2-ethylhexoxy)silane (TEHOS) films on Si wafer with 100 nm thermally grown oxide, and on thin semectic liquid crystal films. Spatial maps of diffusivities from SMT experiments on 220 nm thick semectic liquid crystal films reveal structure related dynamics. The SMT experiments on ultrathin TEHOS films were complemented by fluorescence correlation spectroscopy (FCS). The observed strongly heterogeneous single molecule dynamics within those films can be explained by a three-layer model consisting of (i) dye molecules adsorbed to the substrate, (ii) slowly diffusing molecules in the laterally heterogeneous near-surface region of 1 - 2 molecular diameters, and (iii) freely diffusing dye molecules in the upper region of the film. FCS and SMT experiments reveal a strong influence of substrate heterogeneity on SM dynamics. Thereby chemisorption to substrate surface silanols plays an important role. Vertical mean first passage times (mfpt) in those films are below 1 µs. This appears as fast component in FCS autocorrelation curves, which further contain a contribution from lateral diffusion and from adsorption events. Therefore, the FCS curves are approximated by a tri-component function, which contains an exponential term related to the mfpt, the correlation function for translational diffusion and a stretched exponential term for the broad distribution of adsorption events. Lateral diffusion coefficients obtained by FCS on 10 nm thick TEHOS films, thereby, are effective diffusion coefficients from dye transients in the focal area. They strongly depend on the substrate heterogeneity. Variation of the frame times for the acquisition of SMT experiments in steps of 20 ms from 20 ms to 200 ms revealed a strong dependence of the corresponding probability distributions of diffusivities on time, in particular in the range between 20 ms and 100 ms. This points to average dwell times of the dye molecules in at least one type of the heterogeneous regions (e.g. on and above silanol clusters) in the range of few tens of milliseconds. Furthermore, time series of SM spectra from Nile Red in 25 nm thick poly-n-alkyl-methacrylate (PnAMA) films were studied. In analogy to translational diffusion, spectral diffusion (shifts in energetic positions of SM spectra) can be studied by probability distributions of spectral diffusivities, i.e. time scaled square energetic displacements. Simulations were run and analyzed to study contributions from noise and fitting uncertainty to spectral diffusion. Furthermore the effect of spectral jumps during acquisition of a SM spectrum was investigated. Probability distributions of spectral diffusivites of Nile Red probing vitreous PnAMA films reveal a two-level system. In contrast, such probability distributions obtained from Nile Red within a 25 nm thick poly-n-butylmethacrylate film around glass transition and in the melt state, display larger spectral jumps. Moreover, for longer alkyl side chains a solvent shift to higher energies is observed, which supports the idea of nanophase separation within those polymers.