Several models of inflation with the racetrack superpotential for the volume modulus coupled to a matter field are investigated. In particular, it is shown that two classes of racetrack inflation models, saddle point and inflection point ones, can be constructed in a fully supersymmetric framework with the matter field F-term as a source of supersymmetry breaking and uplifting. Two models of F-term supersymmetry breaking are considered: the Polonyi model and the quantum corrected O'Raifeartaigh model. In the former case, both classes of racetrack inflation models differ significantly from the corresponding models with non-supersymmetric uplifting. The main difference is a quite strong dominance of the inflaton by the matter field. In addition, fine-tuning of the parameters is relaxed as compared to the original racetrack models. In the case of the racetrack inflation models coupled to the O'Raifeartaigh model, the matter field is approximately decoupled from the inflationary dynamics. In all of the above models the gravitino mass is larger than the Hubble scale during inflation. The possibility of having the gravitino much lighter than the Hubble scale is also investigated. It is very hard to construct models with light gravitino in which the volume modulus dominates inflation. On the other hand, models in which the inflationary dynamics is dominated by the matter field are relatively simple and seem to be more natural.

Several models of inflection point inflation with the volume modulus as the inflaton are investigated. Non-perturbative superpotentials containing two gaugino condensation terms or one such term with threshold corrections are considered. It is shown that the gravitino mass may be much smaller than the Hubble scale during inflation if at least one of the non-perturbative terms has a positive exponent. Higher order corrections to the Kahler potential have to be taken into account in such models. Those corrections are used to stabilize the potential in the axion direction in the vicinity of the inflection point. Models with only negative exponents require uplifting and in consequence have the supersymmetry breaking scale higher than the inflation scale. Fine-tuning of parameters and initial conditions is analyzed in some detail for both types of models. It is found that fine-tuning of parameters in models with heavy gravitino is much stronger than in models with light gravitino. It is shown that recently proposed time dependent potentials can provide a solution to the problem of the initial conditions only in models with heavy gravitino. Such potentials can not be used to relax fine tuning of parameters in any model because this would lead to values of the spectral index well outside the experimental bounds.

The relation between the Hubble constant and the scale of supersymmetry breaking is investigated in models of inflation dominated by a string modulus. Usually in this kind of models the gravitino mass is of the same order of magnitude as the Hubble constant which is not desirable from the phenomenological point of view. It is shown that slow-roll saddle point inflation may be compatible with a low scale of supersymmetry breaking only if some corrections to the lowest order Kahler potential are taken into account. However, choosing an appropriate Kahler potential is not enough. There are also conditions for the superpotential, and e.g. the popular racetrack superpotential turns out to be not suitable. A model is proposed in which slow-roll inflation and a light gravitino are compatible. It is based on a superpotential with a triple gaugino condensation and the Kahler potential with the leading string corrections. The problem of fine tuning and experimental constraints are discussed for that model.

It has been conjectured by Ambrosanio, Kane, Kribs, Martin and Mrenna (AKM) that the CDF event $p \bar p \to e^+ e^- \gamma \gamma + missing E_T$ is due to a decay chain involving two neutralino states (the lightest and the next-to-lightest ones). The lightest neutralino ($\chi_{AKM}$) has been further considered by Kane and Wells as a candidate for cold dark matter. In this paper we examine the properties of relic $\chi_{AKM}$'s in their full parameter space, and examine the perspectives for detection by comparing theoretical predictions to sensitivities of various experimental searches. We find that for most regions of the parameter space the detectability of a relic $\chi_{AKM}$ would require quite substantial improvements in current experimental sensitivities. The measurements of neutrino fluxes from the center of the Earth and of an excess of $\bar{p}/p$ in cosmic rays are shown to offer some favorable perspectives for investigating a region of the $\chi_{AKM}$ parameter space around the maximal $\tan \beta$ value allowed by the model.

We examine the properties of relic neutralinos in the regions of the parameter space selected by recent fits to all the electroweak observables within the Minimal Supersymmetric Standard Model. We discuss the relic neutralino cosmic abundance and the direct detection rates for these most likely supersymmetric configurations. We employ the relevant experimental bounds to constrain the set of the best-fit configurations and discuss the discovery potential of the direct search for relic neutralinos.

The allowed parameter space for the lightest neutralino as the dark matter is explored using the Minimal Supersymmetric Standard Model as the low-energy effective theory without further theoretical constraints such as GUT. Selecting values of the parameters which are in agreement with present experimental limits and applying the additional requirement that the lightest neutralino be in a cosmologically interesting range, we give limits on the neutralino mass and composition. A similar analysis is also performed implementing the grand unification constraints. The elastic scattering cross section of the selected neutralinos on $^{27}$Al and on other materials for dark matter experiments is discussed.