This is a doctoral dissertation presented to the FacultyBoard of the Royal Institute of Technology. The dissertationconsists of three self-contained essays on real option pricing.Essay I, written in Swedish, was presented at seminar andaccepted as fulfilling the requirement for a Licentiate Degreein Engineering thesis in 1995. Essay I: This essay studies the option to develop vacantland when the landowner simultaneously determines the optimaldensity and timing of a development. Williams (1991) andCapozza and Li (1994) have recently studied the landdevelopment decision from a realoptions perspective. Bothpapers assume that the production technology is of the commonlyused Cobb-Douglas type, but they use different assumptionsabout uncertainty over future rents and construction costs. Ananalysis of these models and their limitations is carried outand as a result valuation models based on other productiontechnologies than the Cobb-Douglas technology are derived. Essay II: McDonald and Siegel (1996) show that when thebenefit from an investment and the investment cost are assumedto follow correlated geometric Brownian motions, the optimalinvestment policy is given by a simple rule: Invest the firsttime the benefit-cost ratio reaches a certain level. In thisessay, which models the decision to change the use ofaproperty, the investment rule is found to be more complicated.Optimal redevelopment will take place for differentbenefit-cost ratios depending on the relative sizes of thevalue of the property in the different uses and the cost ofchanging the use. Also, for a given current benefit-cost ratiothe value of the option to change use will vary significantlydepending on the relative sizes of the state variables. Essay III: The relationship between the option to choose thecapacity of a real estate development and deliberateoverbuilding is studied in a simple model of investment underuncertainty. The model provides an intuitive measure ofdeliberate overbuilding: the difference between the number ofrental units the owner of an undeveloped site optimally choosesto produce and the number of units expected to be leased at thetime of the building's completion. Numerical simulations withreasonable parameter values show that in some economicenvironments, the optimal production strategy can be to producemore units than are expected to be leased at completion of thedevelopment. / <p>QC 20100611</p>