How to show employees with salaries greater than $1000? In C or C++

Your error is in the function MostarDatosMmil, since the call to printfis fine, but the check is not.

void MostrarDatosMmil()
{
    for (int k = 0; k < i; k++)    // Esto está bien
    {
        if (Cantidad[i].sueldo > sueld)    // Esto está mal
        {
            printf("\n\nNombre: %s         \nApellido: %s         \nEdad: %d          \nSueldo: %0.2f         \nGenero: %s",
                Cantidad[k].nombre, Cantidad[k].apellido, Cantidad[k].edad, Cantidad[k].sueldo, Cantidad[k].genero);
        }
    }
}

You're looping from the first element (0) to the last element (i - 1), but in all iterations you check if the element i (it's after the last one) has a salary less than 1000.

This is going to have undefined behavior, since variables are not initialized by default, and can hold any value, positive, negative, or zero. So sometimes the employee will show you k and sometimes not.

In any case, it's going to be wrong because showing employee k depends on employee i, which hasn't been put into the employee collection yet.

What's more, if you get to fill the vector of employees, it will try to access employee 2001 and you can skip a segment failure for trying to access an area that has not been reserved.

The check ifshould be:

if (Cantidad[k].sueldo > sueld) ...

In any case, consider following C/C++ conventions (you tag your question as C++ but use C libraries, and don't use any C++ features except the new empleados[2000]): write function and variable names starting with lower case ( lower camel case for example), and those #define ALGOin capital letters (without starting with hyphens, since these have a use reserved to the compiler).

Also, I recommend that you don't declare a global variable named ibecause its name doesn't mean anything. Consider refactoring it likenumeroEmpleados for example, and putting full and meaningful variable names.

Also, a type variable struct empleadoswill not contain many employees, but one, so I recommend putting the name in the singular.

For more order and correctness it would be even better to declare another structure, containing the collection (vector) of employees and the number of employees inserted so far.

#define MAXIMO_NUMERO_EMPLEADOS 2000
#define SUELDO_MIL 1000

struct empleado
{
    char nombre[30];
    char apellido[30];
    int edad;
    float sueldo;
    char genero[15];
};

typedef struct empleado Empleado;

struct coleccionEmpleados
{
    Empleado empleados[MAXIMO_NUMERO_EMPLEADOS];
    unsigned numeroEmpleados;
    coleccionEmpleados() : numeroEmpleados(0) {}    // Constructor C++
};

typedef struct coleccionEmpleados ColeccionEmpleados;

ColeccionEmpleados coleccionEmpleados;    // coleccionEmpleados != struct coleccionEmpleados

...

void mostrarDatosMil()
{
    for (int empleado = 0; empleado < coleccionEmpleados.numeroEmpleados; empleado++)
    {
        if (coleccionEmpleados.empleados[empleado].sueldo > SUELDO_MIL) printf(...);
    }
}

You can even define methods within structs to work with them more comfortably and logically. For example:

#define OPCION_MOSTRAR_MIL 3

struct coleccionEmpleados
{
    Empleado empleados[MAXIMO_NUMERO_EMPLEADOS];
    unsigned numeroEmpleados;
    coleccionEmpleados() : numeroEmpleados(0) {}    // Constructor C++
    void mostrarEmpleadosMil()
    {
        for (unsigned empleado = 0; empleado < numeroEmpleados; empleado++) printf(...);
    }
};

typedef struct coleccionEmpleados ColeccionEmpleados;

ColeccionEmpleados coleccionEmpleados;    // Llamada implícita al constructor sin parámetros

int main(int argc, char * argv[])
{
    ...
    switch (opcion)
    {
        ...
        case OPCION_MOSTRAR_MIL:
            coleccionEmpleados.mostrarEmpleadosMil();
            break;
        ...
    }
}

So you don't have free methods that handle structures, instead structures provide methods to handle them properly.